Background/aims: Diagnosis of bile duct strictures as well as detection and localization of hepatolithiasis is important for an effective treatment of hepatolithiasis. For this purpose, endoscopic retrograde cholangiopancreatography (ERCP) and percutaneous transhepatic cholangiography (PTC) have been preferred as the diagnostic modalities of choice for hepatolithiasis. At present, magnetic resonance cholangiopancreatography (MRCP) is rapidly replacing ERCP. To determine the usefulness and limitation of MRCP for detecting hepatolithiasis and biliary strictures, we compared the result of MRCP with that of percutaneous transhepatic cholangioscopy (PTCS) as the standard reference.
Methods: Sixty-six patients with primary hepatolithiasis who underwent both MRCP and PTCS were enrolled. All patients underwent PTCS within 2 weeks of MRCP. The results of MRCP were reviewed by radiologists who were unaware of the clinical information about the location of hepatolithiasis and the presence of biliary strictures.
Results: The sensitivity and specificity of MRCP for detecting intrahepatic stones were 72.4 and 90%, respectively. The sensitivity of MRCP for depicting intrahepatic bile duct strictures was 64%. The overall agreement between MRCP and PTCS for intrahepatic stones showed a good reliability (kappa value=0.64, p<0.05). The agreement of MRCP for hepatolithiasis with intrahepatic bile duct strictures with reference to PTCS showed a moderate reliability (kappa value=0.47, p<0.05). Imaging findings such as pneumobilia and hepatic parenchymal atrophy were related to false positive results in location of hepatolithiasis (p<0.05).
Conclusions: MRCP has a good agreement rate in evaluation of intrahepatic stones and a moderate agreement rate in intrahepatic strictures, compared to PTCS.